Split with single colon but not double colon using regex

Question

I have a string like this

\"yJdz:jkj8h:jkhd::hjkjh\"

I want to split it using colon as a separator, but not a double colon. Desired result:

Answer 1

You could split on (?<!:):(?!:). This uses two negative lookarounds (a lookbehind and a lookahead) which assert that a valid match only has one colon, without a colon before or after it.

To explain the pattern:

(?<!:)  # assert that the previous character is not a colon
:       # match a literal : character
(?!:)   # assert that the next character is not a colon

Both lookarounds are needed, because if there was only the lookbehind, then the regular expression engine would match the first colon in :: (because the previous character isn't a colon), and if there was only the lookahead, the second colon would match (because the next character isn't a colon).

Answer 2

You can do this with lookahead and lookbehind, if you want:

>>> s = "yJdz:jkj8h:jkhd::hjkjh"
>>> l = re.split("(?<!:):(?!:)", s)
>>> print l
['yJdz', 'jkj8h', 'jkhd::hjkjh']

This regex essentially says "match a : that is not followed by a : or preceded by a :"